Mainframe business analysis model training method, business analysis method and device

Abstract

The embodiment of the present application provides a training method, a business analysis method and a device for a mainframe business analysis model, which relates to the field of big data and artificial intelligence technology, including: obtaining multiple training samples, each training sample including a sample business report generated by the mainframe, system status information when the mainframe generates the sample business report, and the annotation result of the sample business report, the annotation result is used to characterize the problem information of the sample business report, and the system status information includes hardware system information and system operation information; according to the multiple training samples, the business analysis model is trained to obtain multiple candidate business analysis models; according to the multiple candidate business analysis models, a target business analysis model is obtained. By analyzing the mainframe's business with a trained mainframe business analysis model, the analysis efficiency can be improved and misjudgment caused by human factors can be reduced.

Description

Technical Field

[0001] The present application relates to the field of big data and artificial intelligence technology, and in particular to a training method, a business analysis method and a device for a mainframe business analysis model. Background Art

[0002] With the rapid development of financial technology, the business volume undertaken by the banking system has also increased significantly. Since mainframes are suitable for handling large-scale business, they have become the core equipment for banks to handle large-scale business.

[0003] Currently, the analysis of possible problems in the mainframe business processing process is mainly carried out manually by experienced analysts.

[0004] However, due to the particularity of manual work, the cost of analysis is high, and manual analysis methods are also prone to missed judgments and misjudgments. Summary of the invention

[0005] The embodiments of the present application provide a mainframe business analysis model training method, a business analysis method and a device. By analyzing the mainframe business with a trained mainframe business analysis model, the analysis efficiency can be improved and misjudgments caused by human factors can be reduced.

[0006] In a first aspect, an embodiment of the present application provides a training method for a mainframe business analysis model, comprising:

[0007] Acquire multiple training samples, each of which includes a sample business report generated by a mainframe, system status information when the mainframe generates the sample business report, and a labeling result of the sample business report, the labeling result is used to characterize problem information of the sample business report, and the system status information includes hardware system information and system operation information;

[0008] Training a business analysis model according to the multiple training samples to obtain multiple candidate business analysis models;

[0009] A target business analysis model is obtained according to the multiple candidate business analysis models.

[0010] Optionally, the business analysis model is trained according to the multiple training samples to obtain multiple candidate business analysis models, including:

[0011] Obtaining M sample sets according to the multiple training samples, where M is an integer greater than or equal to 2;

[0012] The business analysis model is trained according to the M sample sets to obtain M candidate business analysis models.

[0013] Optionally, the training of the business analysis model according to the M sample sets to obtain M candidate business analysis models includes:

[0014] Any sample set among the M sample sets is used as a test set, and the remaining M-1 sample sets are used as training sets;

[0015] For each test set, the business analysis model is trained according to the M-1 training sets to obtain a candidate business analysis model corresponding to the test set;

[0016] According to the candidate business analysis models corresponding to each of the M test sets, the M candidate business analysis models are obtained;

[0017] Optionally, obtaining a target business analysis model according to the multiple candidate business analysis models includes:

[0018] A target business analysis model is obtained according to the M test sets and the candidate business analysis models corresponding to each of the M test sets.

[0019] Optionally, obtaining a target business analysis model according to the multiple candidate business analysis models includes:

[0020] Input each test set into its corresponding candidate business analysis model to obtain M test results;

[0021] According to each of the test results and the annotation results corresponding to each of the test results, the Matthew correlation coefficient corresponding to each of the candidate business analysis models is obtained;

[0022] The target business analysis model is determined according to the Matthew correlation coefficient corresponding to each candidate business analysis model.

[0023] Optionally, determining the target business analysis model according to the Matthew correlation coefficient corresponding to each candidate business analysis model includes:

[0024] According to the Matthew correlation coefficient corresponding to each candidate business analysis model, the maximum Matthew correlation is obtained;

[0025] The candidate business analysis model corresponding to the largest Matthew correlation coefficient is used as the target business analysis model.

[0026] Optionally, obtaining the annotation results of the sample business report includes:

[0027] Sending the sample business report to N user terminals, where N is an integer greater than 2;

[0028] Obtaining N annotation results of the sample business reports fed back by the N user terminals;

[0029] Determine whether the consistency of the N annotation results is greater than a preset value; if so, use the annotation results whose consistency is greater than the preset value as the annotation results of the sample business report.

[0030] Optionally, the hardware system information includes at least one of the following information:

[0031] Processor, memory, coupler, SPOOL peripherals online at the same time, TGS group, network, tape, and number of active jobs;

[0032] The system operation information includes at least one of the following information:

[0033] Processor usage, memory usage, SPOOL usage, and input / output I / O usage.

[0034] In a second aspect, an embodiment of the present application provides a mainframe business analysis method, including:

[0035] Obtaining business reports generated by the mainframe, hardware system information and system operation information when the mainframe generates the business reports;

[0036] The business report, the hardware system information and the system operation information are input into a target business analysis model to obtain an analysis result output by the target business analysis model, wherein the target business analysis model is trained by the method described in any one of the first aspects.

[0037] In a third aspect, an embodiment of the present application provides a training device for a mainframe business analysis model, including:

[0038] An acquisition module, used to acquire multiple training samples, each of which includes a sample business report generated by a mainframe, hardware system information and system operation information when the mainframe generates the sample business report, and a labeling result of the sample business report, wherein the labeling result is used to characterize problem information of the sample business report;

[0039] A training module, used to train the business analysis model according to the multiple training samples to obtain multiple candidate business analysis models;

[0040] The selection module is used to obtain a target business analysis model based on the multiple candidate business analysis models.

[0041] In a fourth aspect, an embodiment of the present application provides a mainframe business analysis device, including:

[0042] An acquisition module, used to acquire the business report generated by the mainframe, the hardware system information and the system operation information when the mainframe generates the business report;

[0043] An analysis module is used to input the business report, the hardware system information and the system operation information into a target business analysis model to obtain an analysis result output by the target business analysis model, wherein the target business analysis model is trained by the method described in any one of the first aspects.

[0044] In a fifth aspect, the present application provides an electronic device, including: a memory and a processor;

[0045] The memory is used to store computer instructions; the processor is used to execute the computer instructions stored in the memory to implement any method in the first aspect or the second aspect.

[0046] In a sixth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the method of any one of the first aspect or the second aspect.

[0047] In a seventh aspect, the present application provides a computer program product, comprising a computer program, which implements the method of any one of the first aspect or the second aspect when executed by a processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A schematic diagram of a scenario provided for an embodiment of the present application;

[0049] Figure 2 Schematic diagram of the training method process of the mainframe business analysis model provided in the embodiment of the present application Figure 1 ;

[0050] Figure 3 Schematic diagram of the training method process of the mainframe business analysis model provided in the embodiment of the present application Figure 2 ;

[0051] Figure 4 A schematic diagram of the mainframe business analysis method provided in the embodiment of the present application;

[0052] Figure 5 A schematic diagram of the structure of a training device for a mainframe business analysis model provided in an embodiment of the present application;

[0053] Figure 6 A schematic diagram of the structure of a mainframe business analysis device provided in an embodiment of the present application;

[0054] Figure 7 A schematic diagram of the structure of an electronic device for training a mainframe business analysis model or performing mainframe business analysis provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to clearly describe the technical solutions of the embodiments of the present application, some terms and technologies involved in the embodiments of the present application are briefly introduced below:

[0056] 1) Mainframes, also known as mainframe computers, use dedicated instruction systems, operating systems, and application software and are primarily used in commercial areas such as banks, telecommunications companies, and large manufacturing companies.

[0057] 2) Matthew correlation coefficient (MCC), a metric used to evaluate classification performance, with an MCC ranging from -1 to 1, where an MCC of -1 indicates the worst possible prediction, and a value of 1 indicates the best possible prediction scenario. In addition, an MCC of 0 indicates a random prediction.

[0058] 3) Other terms

[0059] In the embodiments of the present application, words such as "first" and "second" are used to distinguish the same or similar items with substantially the same functions and effects, and do not limit their order. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit them to be different.

[0060] It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0061] The following is a detailed description of the training method of the mainframe business analysis model provided by the embodiment of the present application in conjunction with the accompanying drawings. It should be noted that the "at..." in the embodiment of the present application can be the instant when a certain situation occurs, or it can be a period of time after a certain situation occurs, and the embodiment of the present application does not specifically limit this.

[0062] With the rapid development of financial technology, the business volume undertaken by the banking system has also increased significantly. Since mainframes are suitable for handling large-scale business, they have become the core equipment for banks to handle large-scale business.

[0063] At present, the analysis of possible problems in the mainframe's business processing process is mainly carried out manually by experienced analysts who refer to the mainframe's system status information.

[0064] However, in the existing process, when analyzing mainframe business, the operation process is lengthy and cumbersome, requiring a large number of operators and operation time, and the cost of analysis is high. In addition, due to the complexity and professionalism of the analysis, analysts can only conduct analysis and evaluation one by one, and the analysis efficiency is low. At the same time, manual analysis methods are also prone to missed judgments and misjudgments.

[0065] In view of this, the present application proposes a training method and a business analysis method for a mainframe business analysis model. By inputting the business report of the mainframe and the corresponding system status information into the trained business analysis model, the analysis results output by the business analysis model are obtained. There is no need to rely on manual analysis, which can improve the analysis efficiency and reduce the errors in the analysis results caused by human factors.

[0066] The following specific embodiments are used to describe in detail the technical solution of the present application and how the technical solution of the present application solves the above technical problems. The following specific embodiments can be implemented independently or in combination with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0067] Figure 1 The schematic diagram of the application scenario of the embodiment of the present application includes a mainframe and a user terminal. The mainframe and the user terminal are connected via a network for data transmission, and a business analysis model for analyzing the mainframe business is provided in the mainframe.

[0068] The mainframe can input the business report that needs to be analyzed and the system information corresponding to the business report into the business analysis model to obtain the analysis results. The mainframe can send the analysis results to the user terminal so that the analysis results can be displayed to the user.

[0069] It is understandable that the mainframe can also use the multiple business reports obtained as sample business reports to train the business analysis model to obtain the expected analysis results. During the training process, the mainframe can annotate the business reports through the user terminal to obtain the annotated results. The mainframe trains the analysis model based on the analysis results and the annotated results.

[0070] The above describes the application scenarios of the embodiments of the present application. Figure 1 Taking the mainframe in as an example, the training method and business analysis method of the mainframe business analysis model provided in the embodiments of the present application are described in detail.

[0071] Figure 2 Schematic diagram of the process of training the mainframe business analysis model provided in the embodiment of the present application Figure 1 , including the following steps:

[0072] S201. Acquire multiple training samples, each training sample including a sample business report generated by a mainframe, system status information when the mainframe generates the sample business report, and a labeling result of the sample business report.

[0073] Business reports are tools used to organize, format, and present data for mainframe business. Business analysts can use business reports combined with mainframe system status information to understand the current mainframe business information and business status in real time. For example, business reports can be used to analyze whether there are problems with mainframe business. Sample business reports are business reports used to train mainframe business analysis models.

[0074] The annotation results of the sample business report are used to characterize the problem information of the business report. They are the analysis results of the sample business report by the user terminal receiving the report on the mainframe based on the hardware system information and system operation information when the mainframe generates the sample business report. For example, the annotation results of the sample business report can be divided into four levels: no problem, minor problem, general problem, and serious problem.

[0075] System status information includes: hardware system information and system operation information.

[0076] Hardware system information refers to the static configuration information of the mainframe hardware system, including but not limited to the processor (central processing unit, CPU), memory, coupler, simultaneous peripheral operations online (simultaneous peripheral operations online, SPOOL), transmission groups (TGS), active job number (active job number, JNUM), network and tape information.

[0077] System operation information refers to the dynamic operation information of the mainframe hardware system, including but not limited to CPU utilization, memory utilization, SPOOL utilization, input / output (I / O) utilization and other information.

[0078] The mainframe uses the generated sample business report, the system status information when the sample business report is generated, and the annotation result of the sample business report as a training sample, and trains the mainframe business analysis model by obtaining multiple training samples.

[0079] S202: Train a business analysis model according to multiple training samples to obtain multiple candidate business analysis models.

[0080] In the embodiment of the present application, the business analysis model can be trained through different training samples to obtain multiple different candidate business analysis models.

[0081] For example, multiple training samples can be divided into multiple different sample sets, and the business analysis model can be trained according to different sample sets by cross-validation to obtain multiple candidate business analysis models. Alternatively, a business analysis model can be trained by one sample set, and the business analysis model can be trained by multiple sample sets to obtain multiple candidate business analysis models.

[0082] When the analysis results output by the business analysis model are inconsistent with the corresponding annotation results, the parameters of the business analysis model can be tuned, and the business analysis model with adjusted parameters can be used for the next round of training until the training is completed.

[0083] S203: Obtain a target business analysis model according to multiple candidate business analysis models.

[0084] After obtaining multiple candidate business analysis models, the multiple candidate business analysis models can be screened to obtain a target business analysis model, which can be a model with the highest matching degree between the analysis results and the annotation results among the multiple candidate business analysis models. It can be understood that the selection method of the target business model can also be set according to actual needs, and the embodiment of the present application does not limit this.

[0085] Exemplarily, the target candidate model can be obtained by calculating multiple candidate model performance evaluation indicators. In the embodiment of the present application, the performance evaluation indicators of the candidate model can be calculated using evaluation methods such as accuracy (AC), sensitivity (SN), specificity (SP) and Matthew correlation coefficient (MCC). It can be understood that other methods can also be used to obtain the target candidate model, and the embodiment of the present application is not limited to this.

[0086] The training method of the business analysis model of the mainframe provided in the embodiment of the present application obtains multiple training samples, each training sample includes a sample business report generated by the mainframe, system status information when the mainframe generates the sample business report, and the annotation result of the sample business report, and trains the business analysis model according to the multiple training samples to obtain multiple candidate business analysis models; and obtains the target business analysis model according to the multiple candidate business analysis models. Using the target business analysis model to analyze the business of the mainframe can effectively reduce the frequency of manual analysis, improve analysis efficiency, and also reduce the error of analysis results caused by human factors.

[0087] Figure 3 Schematic diagram of the process of training the mainframe business analysis model provided in the embodiment of the present application Figure 2 , including the following steps:

[0088] S301. Acquire multiple training samples, each training sample including a sample business report generated by a mainframe and system status information when the mainframe generates the sample business report.

[0089] The implementation method of S301 in the present application embodiment is similar to Figure 2 The implementation method of the illustrated embodiment S201 is similar and will not be repeated here.

[0090] Optionally, after obtaining the sample business report, the sample business report can be formatted for subsequent business analysis model training. Formatting refers to extracting indicator values ​​from the sample business report. The sample business report can also be preliminarily analyzed for subsequent labeling of the sample business report. The analysis methods include but are not limited to descriptive statistics such as mean and median, data visualization analysis such as heat map and principal component analysis, and data shaping analysis. Abnormal data or missing data in the sample business report can also be removed to ensure the effectiveness of subsequent business analysis model training. For example, data with obvious extreme values ​​in the sample business report can be removed.

[0091] Optionally, the mainframe can generate sample business reports with a granularity of one minute using the job control language (JCL) through its own JES2 system, collect the system status information of the mainframe with a granularity of 5 seconds, and match the sample business reports with the corresponding system status information according to the time nodes.

[0092] S302: Obtain the annotation results of each sample business report.

[0093] In the embodiment of the present application, the sample business report may be sent to N user terminals, and N annotation results of the sample business report fed back by the N user terminals may be obtained.

[0094] Determine whether the consistency of the N annotation results is greater than a preset value. If so, take the annotation results whose consistency is greater than the preset value as the annotation results of the sample business report, where N is an integer greater than 2. If not, the sample business report can be sent to another N user terminals, and the process is repeated until the consistency of the N annotation results is greater than the preset value.

[0095] Exemplarily, each report can be distributed to four user terminals, each of which corresponds to an experienced labeler. If the labeling results of three or more labelers are the same, then the labeling result is adopted. Otherwise, the report is distributed to four other different labelers. If the labeling results of three or more labelers are the same, then the labeling result is adopted. Otherwise, it is distributed to four other different labelers again until the labeling results of three or more labelers are consistent. It can be understood that the preset value (the proportion of labelers with the same labeling results to the total number of labelers) can be set according to actual needs. To ensure the rationality of the model training, the proportion of labelers with the same labeling results to the total number of labelers should not be less than 70%.

[0096] S303. Divide the multiple training samples into M sample sets, and train the business analysis model according to the M sample sets to obtain M candidate business analysis models, where M is an integer greater than or equal to 2.

[0097] In the embodiment of the present application, multiple training samples are divided into M sample sets, any sample set in the M sample sets is used as a test set, and the remaining M-1 sample sets are used as training sets.

[0098] For each test set, the business analysis model is trained using the cross-validation method based on M-1 training sets to obtain the candidate business analysis model corresponding to the test set.

[0099] For example, for a selected test set, when using M-1 training sets to train the business analysis model, one training sample in one training set can be first input into the business analysis model. When the analysis result output by the business analysis model is inconsistent with the annotation result, the parameters in the business analysis model are adjusted, and the remaining training samples in the training set are sequentially input into the business analysis model after the parameters are adjusted to obtain the analysis result. The training process is repeated until M-1 training sets are traversed to obtain the candidate business analysis model corresponding to the test set.

[0100] Among the M data sets, select 1 new sample set as the test set, use the remaining M-1 training sets to train the business analysis model, obtain the candidate business analysis model corresponding to the test set, and repeat the training process until all M sample sets are used as test sets to obtain the corresponding M candidate business analysis models.

[0101] S304: Perform performance evaluation on each candidate business selection model.

[0102] In the embodiment of the present application, the Matthew correlation coefficient is used to evaluate the performance indicators of the candidate models.

[0103] Specifically, each test set is input into its corresponding candidate business analysis model to obtain M test results, and the Matthew correlation coefficient corresponding to each candidate business analysis model is obtained according to each test result and the annotation result corresponding to each test result.

[0104] The Matthew correlation coefficient calculation can satisfy the following formula:

[0105]

[0106] Among them, TP is the number of true positive instances, TN is the number of true negative instances, FP is the number of false positive instances, and FN is the number of false negative instances.

[0107] Input multiple training samples in each test set into the corresponding candidate business analysis model in turn to obtain multiple corresponding analysis results. According to the analysis results and the corresponding annotation results, obtain TP, TN, FP, and FN in the test set, and then calculate the Matthew correlation coefficient of the candidate business analysis model corresponding to the test set according to the Matthew correlation coefficient calculation formula.

[0108] The labeling results are divided into four levels: no problem, minor problem, general problem, and serious problem. The analysis results of the test set are also divided into four levels: no problem, minor problem, general problem, and serious problem.

[0109] TP is the number of sample business reports in the test set whose labeled results are one of the following: minor problems, general problems, and serious problems, and whose analysis results are completely consistent with the labeled results. TN is the number of sample business reports whose labeled results are no problems and whose training model prediction results are also no problems; FP is the number of sample business reports whose labeled results are no problems and whose analysis results are one of the following: minor problems, general problems, and serious problems; FN is the number of sample business reports whose labeled results are serious problems and whose analysis results are one of the following: no problems, minor problems, and general problems, or whose labeled results are general problems and whose analysis results are one of the following: no problems, minor problems, or minor problems, or whose labeled results are minor problems and whose analysis results are no problems.

[0110] S305: Determine a target business analysis model according to the Matthew correlation coefficient corresponding to each candidate business analysis model.

[0111] After obtaining the Matthew correlation coefficients corresponding to the M candidate business analysis models, the mainframe obtains the candidate business analysis model corresponding to the maximum Matthew correlation coefficient, and uses the candidate business analysis model with the largest Matthew correlation coefficient as the target business analysis model.

[0112] The training method of the business analysis model of the mainframe provided in the embodiment of the present application obtains multiple training samples and the annotation results of each sample business report in the training samples, divides the multiple training samples into M sample sets, and trains the business analysis model using a cross-validation method to obtain multiple candidate business analysis models; and obtains a target business analysis model based on the Matthew correlation coefficients of the multiple candidate business analysis models. Using the target business analysis model to analyze the business of the mainframe can effectively reduce the frequency of manual analysis, improve analysis efficiency, and also reduce the error of analysis results caused by human factors.

[0113] Figure 4 The flowchart of the mainframe business analysis method provided in the embodiment of the present application includes the following steps:

[0114] S401, obtaining a business report generated by a mainframe, hardware system information and system operation information when the mainframe generates the business report.

[0115] Optionally, the mainframe can generate business reports with JCL at a granularity of one minute through the JES2 system, collect the system status information of the mainframe at a granularity of 5 seconds, and match the business reports with the corresponding system status information according to the time nodes.

[0116] S402: Input the business report, hardware system information and system operation information into the target business analysis model to obtain the analysis result output by the target business analysis model.

[0117] Among them, the target business analysis model is Figure 2 or Figure 3 The model is obtained by training using any of the business analysis model training methods shown.

[0118] Optionally, after obtaining the analysis results output by the target business analysis model, the analysis results corresponding to multiple business reports can be integrated, classified and counted according to the level of the analyzed problems, and a statistical report can be generated. The statistical report can be sent to the user terminal for easy viewing by the user.

[0119] In the embodiment of the present application, by automatically generating business reports on a mainframe, obtaining analysis results corresponding to the business reports and generating reports, the reliance on analysts can be greatly reduced, the analysis result errors caused by human factors can be reduced, and analysis estimates can be made in large quantities, thereby greatly improving evaluation efficiency and reducing test and operation risks.

[0120] Figure 5 The structural diagram of the training device 50 for the mainframe business analysis model provided in the embodiment of the present application includes:

[0121] The acquisition module 501 is used to obtain multiple training samples, each training sample includes a sample business report generated by the mainframe, hardware system information and system operation information when the mainframe generates the sample business report, and the annotation results of the sample business report, and the annotation results are used to characterize the problem information of the business report.

[0122] Optionally, the acquisition module 501 includes a report generation unit, a report formatting unit, a report marking unit, a hardware system information collection unit, and a system operation information collection unit.

[0123] Specifically, the report generation unit is used to generate sample business reports with a granularity of one minute by using JCL through the JES2 system.

[0124] The report formatting unit is used to extract indicator values ​​from the sample business report.

[0125] The report annotation unit is used to send the sample business report to the user terminal and receive the annotation result returned by the user terminal.

[0126] Specifically, a sample business report is sent to N user terminals, and N annotation results of the sample business report fed back by the N user terminals are obtained; it is determined whether the consistency of the N annotation results is greater than a preset value; if so, the annotation results whose consistency is greater than the preset value are used as the annotation results of the sample business report, where N is an integer greater than 2.

[0127] The hardware system information collection unit is used to collect static configuration information of the mainframe hardware system.

[0128] The system operation information collection unit is used for the dynamic operation information of the mainframe hardware system and can collect it at a granularity of 5 seconds.

[0129] The training module 502 is used to train the business analysis model according to multiple training samples to obtain multiple candidate business analysis models.

[0130] Optionally, the training module 502 includes a data exploratory analysis unit, a data cleaning unit, a data cutting unit, and a training unit.

[0131] Specifically, the data exploratory analysis unit is used to conduct preliminary analysis of sample business reports, including descriptive statistics such as mean and median, data visualization analysis such as heat map and principal component analysis, and data shaping analysis.

[0132] The data cleaning unit is used to remove abnormal data or missing data in the sample business report.

[0133] The data cutting unit is used to obtain M sample sets according to a plurality of training samples, and use any sample set in the M sample sets as a test set, and use the remaining M-1 sample sets as training sets.

[0134] The training unit is used to train the business analysis model for each test set according to M-1 training sets to obtain a candidate business analysis model corresponding to the test set; and to obtain M candidate business analysis models according to the candidate business analysis models corresponding to each of the M test sets.

[0135] The selection module 503 is used to obtain a target business analysis model according to multiple candidate business analysis models.

[0136] Specifically, each test set inputs the corresponding candidate business analysis model to obtain M test results. According to each test result and the annotation result corresponding to each test result, the Matthew correlation coefficient corresponding to each candidate business analysis model is obtained. According to the Matthew correlation coefficient corresponding to each candidate business analysis model, the maximum Matthew correlation is obtained, and the candidate business analysis model corresponding to the largest Matthew correlation coefficient is used as the target business analysis model.

[0137] The training device for the mainframe business analysis model provided in the embodiment of the present application can execute Figure 2 and / or Figure 3 The technical solution of the method embodiment shown has similar implementation principles and technical effects, which will not be repeated here.

[0138] Figure 6 The structural diagram of the mainframe business analysis device 60 provided in the embodiment of the present application includes:

[0139] Acquisition module 601, for acquiring business reports generated by the mainframe, hardware system information and system operation information when the mainframe generates business reports.

[0140] The analysis module 602 is used to input the business report, hardware system information and system operation information into the target business analysis model to obtain the analysis result output by the target business analysis model.

[0141] Among them, the target business analysis model is Figure 2 or Figure 3 The model is obtained by training using any of the business analysis model training methods shown.

[0142] The mainframe business analysis device provided in the embodiment of the present application can be executed Figure 4 The technical solution of the method embodiment shown has similar implementation principles and technical effects, which will not be repeated here.

[0143] Figure 7This is a schematic diagram of the structure of the mainframe business analysis model training or mainframe business analysis electronic device provided in the embodiment of the present application. Figure 7 As shown, the mainframe business analysis model training or mainframe business analysis electronic device 70 provided in this embodiment may include:

[0144] Processor 701.

[0145] The memory 702 is used to store executable instructions of the terminal device.

[0146] Among them, the processor is configured to execute the above-mentioned mainframe business analysis model training method or the technical solution of the mainframe business analysis method embodiment by executing executable instructions. Its implementation principle and technical effect are similar and will not be repeated here.

[0147] A computer-readable storage medium is also provided in an embodiment of the present application, on which a computer program is stored. When the computer program is executed by a processor, the training method of the mainframe business analysis model or the technical solution of the mainframe business analysis method embodiment is implemented. The implementation principle and technical effect are similar and will not be repeated here.

[0148] In one possible implementation, a computer-readable medium may include a random access memory (RAM), a read-only memory (ROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, a magnetic disk storage or other magnetic storage device, or any other medium that is intended to carry or store the required program code in the form of instructions or data structures and can be accessed by a computer. Moreover, any connection is appropriately referred to as a computer-readable medium. For example, if the software is transmitted from a website, server or other remote source using a coaxial cable, a fiber optic cable, a twisted pair, a digital subscriber line (DSL) or wireless technologies such as infrared, radio and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL or wireless technologies such as infrared, radio and microwave are included in the definition of the medium. Disks and optical disks as used herein include optical disks, laser disks, optical disks, digital versatile disks (DVDs), floppy disks and Blu-ray disks, where disks usually reproduce data magnetically, while optical disks reproduce data optically using lasers. Combinations of the above should also be included within the scope of computer-readable media.

[0149] A computer program product is also provided in an embodiment of the present application, including a computer program. When the computer program is executed by a processor, it implements the technical solution of the training method of the mainframe business analysis model and / or the mainframe business analysis method embodiment. The implementation principle and technical effects are similar and will not be repeated here.

[0150] In the specific implementation of the above-mentioned terminal device or server, it should be understood that the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. The general-purpose processor can be a microprocessor or the processor can be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware processor, or can be executed by a combination of hardware and software modules in the processor.

[0151] Those skilled in the art will appreciate that all or part of the steps of any of the above method embodiments may be completed by hardware associated with program instructions. The aforementioned program may be stored in a computer-readable storage medium, and when the program is executed, all or part of the steps of the above method embodiments are executed.

[0152] If the technical solution of the present application is implemented in the form of software and sold or used as a product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the technical solution of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a computer program or several instructions. The computer software product enables a computer device (which can be a personal computer, a server, a network device, or a similar electronic device) to perform all or part of the steps of the method of the embodiment of the present application.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A training method for a mainframe business analysis model, characterized in that: include: Acquire multiple training samples, each of which includes a sample business report generated by a mainframe, system status information when the mainframe generates the sample business report, and a labeling result of the sample business report, the labeling result is used to characterize problem information of the sample business report, and the system status information includes hardware system information and system operation information; Obtaining M sample sets according to the multiple training samples, where M is an integer greater than or equal to 2; Any sample set among the M sample sets is used as a test set, and the remaining M-1 sample sets are used as training sets; For each test set, the business analysis model is trained according to M-1 training sets to obtain a candidate business analysis model corresponding to the test set; According to the candidate business analysis models corresponding to the M test sets, M candidate business analysis models are obtained; Input each test set into its corresponding candidate business analysis model to obtain M test results; According to each of the test results and the annotation results corresponding to each of the test results, the Matthew correlation coefficient corresponding to each of the candidate business analysis models is obtained; According to the Matthew correlation coefficient corresponding to each candidate business analysis model, the maximum Matthew correlation coefficient is obtained; The candidate business analysis model corresponding to the largest Matthew correlation coefficient is used as the target business analysis model; For each test set, the business analysis model is trained according to M-1 training sets to obtain a candidate business analysis model corresponding to the test set, including: Input a training sample from a training set into the business analysis model. When the analysis result output by the business analysis model is inconsistent with the annotation result, adjust the parameters in the business analysis model, and input the remaining training samples in the training set into the business analysis model after the adjustment of the parameters in turn to obtain the analysis result. Repeat the training process until the M-1 training sets are traversed to obtain the candidate business analysis model corresponding to the test set. The labeling results of the sample business report are obtained, including: Sending the sample business report to N user terminals, where N is an integer greater than 2; Obtaining N annotation results of the sample business reports fed back by the N user terminals; Determine whether the consistency of the N annotation results is greater than a preset value; if so, use the annotation results whose consistency is greater than the preset value as the annotation results of the sample business report.

2. The method according to claim 1, characterized in that The hardware system information includes at least one of the following information: Processor, memory, coupler, SPOOL peripherals online at the same time, TGS group, network, tape, and number of active jobs; The system operation information includes at least one of the following information: Processor usage, memory usage, SPOOL usage, and input / output I / O usage.

3. A mainframe business analysis method, characterized in that: include: Obtaining business reports generated by the mainframe, hardware system information and system operation information when the mainframe generates the business reports; The business report, the hardware system information and the system operation information are input into a target business analysis model to obtain an analysis result output by the target business analysis model, wherein the target business analysis model is trained by the method according to claim 1 or 2.

4. A training device for a mainframe business analysis model, characterized in that: include: An acquisition module, used to acquire multiple training samples, each of which includes a sample business report generated by a mainframe, hardware system information and system operation information when the mainframe generates the sample business report, and a labeling result of the sample business report, wherein the labeling result is used to characterize problem information of the sample business report; A training module, used to train the business analysis model according to the multiple training samples to obtain multiple candidate business analysis models; A selection module, configured to obtain a target business analysis model based on the multiple candidate business analysis models; The training module is specifically used to obtain M sample sets according to the multiple training samples, where M is an integer greater than or equal to 2; any sample set in the M sample sets is used as a test set, and the remaining M-1 sample sets are used as training sets; For each test set, the business analysis model is trained according to M-1 training sets to obtain a candidate business analysis model corresponding to the test set; according to the candidate business analysis models corresponding to each of the M test sets, the M candidate business analysis models are obtained; The selection module is specifically used to input each test set into the corresponding candidate business analysis model to obtain M test results; according to each of the test results and the annotation results corresponding to each test result, obtain the Matthew correlation coefficient corresponding to each candidate business analysis model; According to the Matthew correlation coefficient corresponding to each candidate business analysis model, the maximum Matthew correlation is obtained; and the candidate business analysis model corresponding to the maximum Matthew correlation coefficient is used as the target business analysis model; The training module is specifically used to input a training sample in a training set into the business analysis model, and when the analysis result output by the business analysis model is inconsistent with the annotation result, adjust the parameters in the business analysis model, and input the remaining training samples in the training set into the business analysis model after the adjustment of the parameters in turn to obtain the analysis result; repeat the training process until the M-1 training sets are traversed to obtain the candidate business analysis model corresponding to the test set; The acquisition module is specifically used to send the sample business report to N user terminals, where N is an integer greater than 2; obtain N annotation results of the sample business report fed back by the N user terminals; determine whether the consistency of the N annotation results is greater than a preset value; if so, use the annotation results whose consistency is greater than the preset value as the annotation results of the sample business report.

5. A mainframe business analysis device, characterized in that: include: An acquisition module, used to acquire the business report generated by the mainframe, the hardware system information and the system operation information when the mainframe generates the business report; An analysis module is used to input the business report, the hardware system information and the system operation information into a target business analysis model to obtain an analysis result output by the target business analysis model, wherein the target business analysis model is trained by the method described in claim 1 or 2.

6. An electronic device, characterized in that: include: Memory for storing computer programs; A processor, configured to execute the computer program to implement the method of claim 1 or 2, or to implement the method of claim 3.

7. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program is executed by a processor to implement the method of claim 1 or 2, or to implement the method of claim 3.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, it implements the method according to claim 1 or 2, or it implements the method according to claim 3.

Images